This is a competitive renewal of the NIH funded study "Methods of Predicting Delayed Cerebral Ischemia (DCI) in subarachnoid hemorrhage (SAH)". In our current study, we found that 33% of SAH patients develop symptomatic vasospasm (SV) and DCI, and have nearly twice the mortality rate than patients that do not develop SV. We found the sensitivity and specificity of alternative, noninvasive techniques to identify SV and DCI in this population disappointing. We now wish to extend our work and examine a biomarker specific for vasospasm, (20-HETE), that has not been tested in humans. Several well known vasoactive agents [norepinephrine (NE), nitric oxide (NO), endothelin (ET-1)] depend on the formation of 20-HETE to elicit the majority of their vasoconstrictive effects. Independently these vasoactive agents have not proven to be sufficiently sensitive for early detection of SV, but their cumulative effect through 20-HETE formation hold promise as a clinical indicator of SV and DCI. The specific aims are to: 1) Characterize the individual recovery curves of norepinephrine (NE), nitric oxide (NO), endothelin (ET-1), and 20-HETE in the CSF within the first 14 days after SAH., 2) Define the time dependent relationship between 20-HETE and outcomes [acute complications (SV, DCI) and function 3 months post SAH], and 3) Define the relationship between NE, NO and ET-1 in the CSF and outcomes. This study will use a prospective, longitudinal, within-subject between-group repeated measure design to compare 20-HETE, NE, NO, and ET-1 (during the first 14 days following SAH) between individuals who do and do not develop SV or DCI. We will recruit 180 patients with SAH (ages 21-75 years) in order to achieve a final sample of 150 subjects available at the three-month follow-up. All subjects will undergo serial sampling of cerebrospinal fluid, serum, and urine samples, concurrent with intense neurophysiologic monitoring and clinical examinations, to detect early markers of SV and DCI. The independent variables are the CSF levels of NE, NO, ET-1 and 20-HETE. The acute dependent variable outcomes are: SV, DCI and functional outcomes at 3 months post hemorrhage. Hierarchical linear modeling (HLM) (i.e., multilevel modeling or growth curve modeling) analysis will be used to examine the recovery curves of 20-HETE. NE, NO, and ET-1.